The Dragonfly Telephoto Array


Forrest Erickson
 

Facebook alerted me to this Telephoto Array.
 
They are doing interesting things with their array. They describe the system as
1.0 m diameter refractor with a focal ratio of f/0.4.

I have sent an email to a principal investigator to ask for more technical information. 
 
Lee


Forrest Erickson
 

The contact got back to me right away with this:
 
Hi Forrest,
The current cameras on the back are SBIG STT-8300Ms but we’re in middle of a big upgrade (going from 48 to 168 lenses, and from 2 arrays to 6) and in this process we are upgrading all the cameras to be SBIG ALUM 694s. We’re also adding a tunable narrow band imaging capability.
 
Everything is being read out using some bespoke software that I wrote. Each camera is controlled by a node.js webserver running on an Intel compute stick (one per lens) centrally coordinated by a bunch of Python and Perl scripts running on a master control computer. Once the images are taken everything gets pushed to the cloud, where they get stacked by a bespoke AstroPy-based pipeline running on a bunch of virtual machines. The pipeline was written by our amazing grad students.
 
Hope this helps!
Bob
 
I have asked if the software will be published.
Lee

On 05/14/2021 10:06 AM Forrest Erickson <forresterickson@...> wrote:
 
 
Facebook alerted me to this Telephoto Array.
 
They are doing interesting things with their array. They describe the system as
1.0 m diameter refractor with a focal ratio of f/0.4.

I have sent an email to a principal investigator to ask for more technical information. 
 
Lee


Forrest Erickson
 

I asked some questions (below) and got back this interesting information which of course is hard to read because it is in reverse order.
______________________________________________________
 
 
Hi Forrest,
 
We gather the data 48x faster and that’s important, but it’s just part of the reason the system is designed the way it is. Field pointings in each lens are displaced from each other by about 20% of the field diameter, and since we take all the data simultaneously, we are able to average over sources of systematic error (which are sometimes internal, such scattered light and ghosts from the optics, and external, such as variations in the wide-angle PSF due to aerosols, etc.). These systematics dominate at low surface brightness levels and taming them is a key aspect of the survey design.
 
This paper shows the current system and I think it will answer all your questions:
 
 
Dragonfly is evolving really quickly so the details are likely to change soon, but that paper is an accurate description of the system as of right now.
 
At some point we will definitely be making all our software publicly available. I think that’s at least a year away though. The system is behind firewalls but before we release our code (and make all our raw data available, which is part of the plan) the students need to write a few more papers and then I want to do a network security audit of everything.
 
Hope this helps,
Bob
--
Prof. Roberto Abraham, FRSC
Chair, David A. Dunlap Department of Astronomy & Astrophysics University of Toronto
50 St. George Street
Toronto, Ontario M5S 3H4
CANADA
Pronouns: he, him, his
Phone (my office): 416-946-7289
Phone (general enquiries): 416-946-5243
On the web:
 
 
 

From: FORREST ERICKSON < forresterickson@...>
Sent: May 14, 2021 10:21 AM
To: Roberto Abraham < roberto.abraham@...>
Subject: RE: Dragonfly Optical Description
EXTERNAL EMAIL:
 
Hello Bob,
Thanks for the summary. 
Initially from the head line I though, Are they combining the images from lenses on to one sensor for the image resolution of 1M?, but when I saw the photographs of the array I saw what I assume are the STT-833Ms.
 
So you have a system that does 48 images in parallel for "parallel" stacking.   So you gather data 48 time faster than I would with a single STT-833M taking images serially.

Are you and the students open publishing the software for the system?
I am sure some amateurs would be interested in building two camera systems and working their way up.
 
And if any students want to present their work to hobbyist I think we would be a receptive audience. 
 
Thanks,
Forrest Erickson
Phone 865-518-1040
Maryville TN
 
On 05/14/2021 9:52 AM Roberto Abraham < roberto.abraham@...> wrote:
 
Hi Forrest,
The current cameras on the back are SBIG STT-8300Ms but we’re in middle of a big upgrade (going from 48 to 168 lenses, and from 2 arrays to 6) and in this process we are upgrading all the cameras to be SBIG ALUM 694s. We’re also adding a tunable narrow band imaging capability.
 
Everything is being read out using some bespoke software that I wrote. Each camera is controlled by a node.js webserver running on an Intel compute stick (one per lens) centrally coordinated by a bunch of Python and Perl scripts running on a master control computer. Once the images are taken everything gets pushed to the cloud, where they get stacked by a bespoke AstroPy-based pipeline running on a bunch of virtual machines. The pipeline was written by our amazing grad students.
 
Hope this helps!
Bob
--
Prof. Roberto Abraham, FRSC
Chair, David A. Dunlap Department of Astronomy & Astrophysics University of Toronto
50 St. George Street
Toronto, Ontario M5S 3H4
CANADA
Pronouns: he, him, his
Phone (my office): 416-946-7289
Phone (general enquiries): 416-946-5243
On the web:
 
From: FORREST ERICKSON < forresterickson@...>
Sent: May 14, 2021 9:44 AM
To:  abraham@...
Cc: Duane Dunlap < lddunlap41@...>; James Cantu < james.cantu@...>
Subject: Dragonfly Optical Description
 
 
 
Hello Dr. Abraham,
I am an amateur astronomy and member of a club in east Tennessee, the Smoky Mountain Astronomical Society. 
I recently read about the Dragonfly system and found the web site at: 
I read about the Dragonfly research publications but could not find any  detailed description about the optical and image sensor systems. From the photos of the array I think I see cameras on each lens which has me curious on how the images are captured and combined.
 
Is there presentation with some technical details available?  
 
Many of our members are experienced amateur images and would benefit I am sure.  As someone with 100mm F5, I recognize that the Dragonfly specification as  1.0 m diameter refractor with a focal ratio of f/0.4, is impressive and so will other members of our club. 
 
Thanks,
Forrest Erickson
Maryville TN
 
cc Officers of the Smoky Mountain Astronomical Society
 
On 05/14/2021 10:06 AM Forrest Erickson <forresterickson@...> wrote:
 
 
Facebook alerted me to this Telephoto Array.
 
They are doing interesting things with their array. They describe the system as
1.0 m diameter refractor with a focal ratio of f/0.4.

I have sent an email to a principal investigator to ask for more technical information. 
 
Lee